The data were already captured; the question was how to read them

There is a particular anxiety that follows a breast cancer diagnosis, a question that lingers beneath every follow-up appointment and every scan: will it come back? For women with the most common subtype of the disease—hormone receptor-positive, HER2-negative breast cancer—the answer has traditionally depended on a combination of pathology, clinical factors, and genomic tests that can take weeks to return. A new tool, cleared by the Food and Drug Administration this month, offers a different approach, one that draws on information already captured in a patient's MRI to generate a recurrence risk score before surgery even happens.

The tool is called TumorSight Risk, developed by the Chicago-based company SimBioSys. It received 510(k) clearance from the FDA in September and was announced publicly in October. The software uses breast MRI data, along with pathology and clinical information, to estimate the risk of recurrence at five and ten years for women with Stage I to IIIA hormone receptor-positive, HER2-negative breast cancer. It produces a score from 0 to 100, categorizing patients as low, intermediate, or high risk .

The validation data, presented at the 2026 ASCO Annual Meeting, came from a retrospective study of 2,129 patients across four institutions. Patients classified as high risk had a 3.3-fold higher risk of recurrence than those classified as low risk. Each ten-point increase in the score corresponded to a 22 percent increase in relative recurrence risk. Five-year recurrence-free survival was 95.5 percent for low-risk patients and 89.6 percent for high-risk patients; at ten years, the figures were 88.9 percent and 69.6 percent . The association remained significant after adjusting for tumor stage, grade, age, race, and treatment.

What distinguishes TumorSight Risk from existing genomic assays is not necessarily superior accuracy but accessibility and timing. The tool uses data already collected during routine care. It requires no additional procedure, no tissue shipment, no specialized hardware. Results can be available within standard clinical workflows, potentially before surgery, when treatment decisions are being weighed. The company describes it as complementary to existing methods, not a replacement .

The clinical implications are still being worked out. The score is prognostic—it estimates risk—but it does not predict who will benefit from chemotherapy or other treatments. Whether it improves outcomes, reduces unnecessary therapy, or expands access to personalized risk assessment are questions that only clinical use can answer. The FDA clearance permits marketing; it does not constitute proof of clinical utility .

There is a quiet significance in the shift this represents. For decades, the information that shaped cancer treatment came from tissue—what the pathologist saw under the microscope, what the genomic assay read in the tumor's DNA. TumorSight Risk draws from imaging, from the spatial relationships between tumor and tissue that MRI captures. It is a different kind of data, and it is available earlier. Whether that difference matters will be determined by the oncologists and patients who use it, and by the outcomes that follow.

For now, the tool is available, the validation data are published, and the question of how it fits into the crowded landscape of breast cancer prognostication is open. The score is a number, and numbers can be clarifying or misleading. What the score offers is a conversation starter, a piece of information that a woman and her doctor can hold alongside everything else they know. The anxiety will not disappear. But the information may arrive sooner, and that is not nothing.

---

AI Image Disclaimer: The visuals in this article are generated by artificial intelligence and are intended for illustrative purposes only.

Sources: ASCO AI, Fierce Biotech, MedPath Trial, FDA 510(k) database